arXiv subjects

Madhusudan Madhavan

Publications and source records attributed to Madhusudan Madhavan.

3 recordsLinked to original sources

arXiv2026

Hyper-differential sensitivity analysis with respect to model discrepancy: Sequential optimal experimental design

Large-scale optimization problems are ubiquitous in the physical sciences; yet, high-fidelity models can often be complex and computationally prohibitive for optimization. A practical alternative is to use a low-fidelity model to facilitate optimization. However, the discrepancy between the high- and low-fidelity models can lead to suboptimal solutions. To address this, we build on recent work in Hyper-Differential Sensitivity Analysis to leverage limited high-fidelity simulations to update the optimization solution. Our contributions in this article include: (i) incorporating pseudo-time continuation techniques to efficiently compute higher-accuracy optimal solution updates, and (ii) proposing a Bayesian framework for sequential data acquisition that strategically guides high-fidelity evaluations and reduces uncertainty in the model discrepancy estimation. Numerical results demonstrate that our framework delivers significant improvements to optimization solutions with only a few high-fidelity evaluations.

math.NA↗

arXiv2026

FlexTrace: Exchangeable Randomized Trace Estimation for Matrix Functions

We consider the task of estimating the trace of a matrix function, ${\rm tr}(f({\bf A}))$, of a large symmetric positive semi-definite matrix ${\bf A}$. This problem arises in multiple applications, including kernel methods and inverse problems. A key challenge across existing trace estimation methods is the need for matrix-vector products (matvecs) with $f({\bf A})$, which can be very expensive. In this article, we introduce a novel trace estimator, FlexTrace, an exchangeable, single-pass method that estimates ${\rm tr}(f({\bf A}))$ solely using matvecs with ${\bf A}$. We consider the case where $f$ is an operator monotone matrix function with $f(0)=0$, which includes functions such as $\log(1+x)$ and $x^{1/2}$, and derive probabilistic bounds showcasing the theoretical advantages of FlexTrace. Numerical experiments across synthetic examples and application domains demonstrate that FlexTrace provides substantially more accurate estimates of the trace of $f({\bf A})$ compared to existing methods.

math.NA↗

arXiv2025

A control-oriented approach to optimal sensor placement

We propose a control-oriented optimal experimental design (cOED) approach for linear PDE-constrained Bayesian inverse problems. In particular, we consider optimal control problems with uncertain parameters that need to be estimated by solving an inverse problem, which in turn requires measurement data. We consider the case where data is collected at a set of sensors. While classical Bayesian OED techniques provide experimental designs (sensor placements) that minimize the posterior uncertainty in the inversion parameter, these designs are not tailored to the demands of the optimal control problem. In the present control-oriented setting, we prioritize the designs that minimize the uncertainty in the state variable being controlled or the control objective. We propose a mathematical framework for uncertainty quantification and cOED for parameterized PDE-constrained optimal control problems with linear dependence to the control variable and the inversion parameter. We also present scalable computational methods for computing control-oriented sensor placements and for quantifying the uncertainty in the control objective. Additionally, we present illustrative numerical results in the context of a model problem motivated by heat transfer applications.

math.OC↗